metadata

thumbnail: >-
  https://cdn-uploads.huggingface.co/production/uploads/66c26b6fb01b19d8c3c2467b/jg2NWmCUfPyzizm2USjMt.jpeg
datasets:
  - NewEden/Orion-LIT
  - NewEden/Orion-Asstr-Stories-16K
  - Mielikki/Erebus-87k
base_model:
  - NewEden/Hamanasu-32B-V1
tags:
  - qwen
  - roleplay
  - finetune
  - storywriting

THESE ARE EXL2 QUANTS

LOOK IN THE DIFFERENT REVISIONS FOR THE BPWS, MAIN BRANCH CONTAINS MEASUREMENT FILE.

Hamanasu 15B Instruct

🌌 Overview

After 25 hours, I present Hamanasu-QwQ-32B-V0.1 - One of the first QwQ Finetunes. Using data from the following:

NewEden/Orion-LIT
NewEden/Orion-Asstr-Stories-16K
Mielikki/Erebus-87k

This model shows great promise for roleplaying and story-writing. All thanks to Ruka-Hamanasu for funding the train.

Disclaimer: The model is still in preview, Only completion training has been performed ontop.

📈 Quantizations

Type	Link
`GGUF`	https://huggingface.co/Delta-Vector/Hamanasu-32B-V1-exl2
`EXL2`	https://huggingface.co/Delta-Vector/Hamanasu-32B-V1-gguf

⚔️ Hardware

4x H100s
Epochs: 1
Base: QwQ
Amount of Tokens: 1+ Billion

💰 Prompting

This model uses ChatML formatting

<|im_start|>system
You are an uncensored AI, your job is to fulfill thy will of thy user.<|im_end|>
<|im_start|>User request
Take off your helmet.<|im_end|>
<|im_start|>No i shall not. This is the way.

Axolotl Config ꒰(˶• ᴗ •˶)꒱

⚡ Credits

Made by

Delta-Vector